For specific issues, I like to use examples as much as I can and they seem to work really well. I started a small collection of research, real-life examples for the very same reason at www.defendyourdesign.com.

Also, I like to prepare all documentation (from concepts to wireframes) with documenting both how I got to the solution and how it will support the goals that we have. This also seems to smooth discussions on designs.

But if there are usability findings for the project at hand, that works the best. Like a charm almost always.

There's an interesting article published today on A List Apart called "Can You Say That in English? Explaining UX Research to Clients" that outlines some nice cost benefits, and ways to communicate them, for UX research

http://www.alistapart.com/articles/can-you-say-that-in-english-explaining-ux-research-to-clients/

HTH

Alex

Personally, I always find talking about money (to the management) gets them interested quickly.

Show them how you can actually save/make money if you take care of the site's usability/user experience. You could even take it a step further and estimate how much money you'd save/make through the site if it was easier and more pleasant to use. Of course, you need to base your claims on real usage data.

People might think UX/usability is interesting, but they won't really invest in it until it actually save/make them money, and it does most of the time.

Hope this helps :)

Create an initial implementation or "proof of concept" of your design, and let end users play with it. We've done that, and their reactions – "That's so much better!" "Oh, this is going to save me so much time." "I love this!" – make a compelling case for the defence. It's not always easy to set up, but we've found it to be very effective.

I don't know why but you can easily impress stakeholders by showing them a great number of early sketches and mockups. Even if they understand nothing at all, sketches may work like magic. But though it's invaluable to impress it doesn't work so well with little details and particular design changes.

I use the user stats and error logs to do some quick flashy charts in a power point that they can shoot up the chain. I employ some sketches/wireframes that show how we could easily decrease some of the errors we see or increase access to pages users aren't visiting.

Quite frankly I only use it for management, it's not really good/complete analysis and sketches/wireframes are always just "ideas", not what we actually end up implementing after real research.

"Institutionalization of Usability": A Step-by-Step Guide by Eric Schaffer contains quite a lot of useful content regarding your question. Recommended.